Скачать или смотреть Building Data-driven User Persona's using Unsupervised Learning - Jade Abbott & Chantelle Withfield

We are not our users; we probably think that we know what they want, but we rarely do. The field of UX aims to understand and represent users's needs, and often builds "user personas" to do so. These user persona's provide detailed descriptions about the types of users - but these too are problematic as they capture the biases, and assumptions of the designer and fall suspectible to their blindspots.

By running customer surveys and employing unsupervised machine learning, we gained insight into the clusters of our users, we were empowered to create more intelligent, data-driven personas.

This session will explain what User Persona's are, what Unsupervised Learning is, and how the two were brought together to create a fuller view of our users. We'll also delve into the many mistakes and learnings along the way - there are many things we'd do differently next time